New Trends and Prospects for the Distribution of International Taxation Rights with looking at the tradition since the era of the League of NationsASATSUMA Akiyuki

I. Outline of this article

＊1

This article tries to look at the image of source of income (source has two meanings;

{from what} and {from where} which is called as geographical allocation of income in this article) in the traditional international tax law system since the era of the League of Nations, in order to deeply understand contemporary change of international taxing rights allocation (especially allocated to a place of demand).

New Trends and Prospects for the Distribution of International Taxation Rights with looking at the tradition since the era of the League of Nations ASATSUMA Akiyuki

Professor, College of Law and Politics, Rikkyo University, Tokyo, Japan.

Abstract

The most important choice in the era of the League of Nations is the adoption of sepa- rate accounting, which considers a domestic physical presence (which corresponds to the present concept of a “permanent establishment” (PE)) of a foreign enterprise to be an inde- pendent enterprise. The idea of separate accounting , in which a person (individual or corpo- ration) or a part (PE) that physically contributes to the business income is entitled to a share of the income according to its contribution, is associated with the idea of the labor value theory. In reality, entities sometimes do receive income without contribution, such as in compensation for covenants not to compete. When we allocate income among affiliated en- terprises in accordance with the contributions made, there are two types of independent business transactions, or arm’s length transactions, of note: reliable arm’s length transac- tions, and unreliable arm’s length transactions.

We can find examples of both mild derogation from the arm’s length principle and se- vere derogation from the arm’s length principle. In one case, Amount B, in which profit is attributable to a place of activities according to fictious rate of return, can be considered as mild derogation from the arm’s length principle, in which transactions like unreliable arm’s length transactions are denied. In another case, Amount A, in which taxing rights are allo- cated to places of demand, can in no way be justified as mild derogation from the arm’s length principle, and must be considered as severe derogation from the arm’s length princi- ple.

Keywords: international tax law, League of Nations, source rule, permanent establish- ment, threshold, separate accounting, arm’s length principle

JEL Classification: H25, H26

＊1

In this article, no job title or honorific title is added to a person’s name. “ ” or ‘ ’ is used for quotations, and { } is used for clarification. YYYYMMDD is used, and MMDDYYYY or DDMMYYYY is used only in quotations.

This article is originally written in Japanese in Financial Review, no. 143 (2020 June).

Chapter II reviews the framework of the existing international tax law system.

Section II-1 reviews source of income rules.

Source of income rules have standards for earners (mainly for sales income, service in- come or other business income) and standards for payers (for dividends income, interest in- come, royalty income, or etc.). If we consider that source of income rules for dividends in- come and interest income looks at the place of producing activities of payers, most source of income rules can be consistently understood as according to standard of place of production (although there are differences between place of earners’ production and place of payers’

production). Nevertheless, because it is difficult to justify source rule of royalty income (due to differences in source rules of dividends income and interest income) looking at place of payers’ producing activities, we must recognize that source rule of royalty income is dero- gated from a consistent understanding of other source of income rules looking at place of production.

Section II-2 discusses the important choices made in the era of the League of Nations.

In the light of constraints of tax enforcement capacity in the era of the League of Na- tions, when a country (Country S) tried to impose tax on an enterprise of a foreign country (Country R), there was little space of alternatives to require a domestic physical presence (which corresponds to the present concept of a “permanent establishment” (PE)) of the en- terprise. In other words, the {no taxation without a PE} rule was not a result of choices made in the era of the League of Nations.

An important choice made in the era of the League of Nations was the adoption of the

idea of separate accounting, in which a physical presence in Country S of an enterprise of

Country R is considered as an independent enterprise, or the adoption of the idea of consoli-

dation, in which the presence in Country S is considered as a part of international expansion

of the enterprise of Country R. Even in that era, there were some thoughts that the idea of

separate accounting does not fit with the practical reality of an enterprise’s international ex-

pansion (the practical reality in which the enterprise creates an alter ego because the enter-

prise expects to make more profit when creating the alter ego than entering into market

transactions with third parties); however, constraints of tax enforcement capacity of Country

S of that era had difficulty in acquiring information in Country R in order to enforce tax ac-

cording to the idea of consolidation, and lead to the adoption of the idea of separate account-

ing. But, even if we assume the constraints of tax enforcement capacity of that era, if we do

not stick with separate accounting and if we have an image, not of the {no taxation without

a PE} rule, but of the {no taxation without a PE or an affiliated entity} rule, Country S could

have applied the {force of attraction} rule (like the entire income principle in Japan before

the 2014 amendment, all Country S’ domestic source income, regardless whether attribut-

able to a PE or not, is included into taxable income in Country S in the procedure of tax fil-

ing by the PE), not only to a PE of the enterprise of Country R which has only one legal

personality, but also to an affiliated entity of the enterprise of Country R without regard to

the number of legal personalities. Not only when Country R’s enterprise (X co.) has a physi-

cal presence as a PE in Country S, but also when X co.’s affiliated entity (X co.’s affiliated

corporation established in Country S or a branch established in Country S of X co.’s affiliat- ed corporation not established in Country S) is physically located in Country S, Country S can impose obligation of tax filing on the physical presence in Country S (X co.’s branch, X co.’s affiliated co., or a branch of X co.’s affiliated co.) concerning the income which can be considered as Country S’ domestic source income by investigations engaged in Country S by Country S’ tax office, regardless whether the income is legally attributable to X co.’s head office in Country R, X co.’s branch in Country S, X co.’s affiliated co. in Country S, X co.’s affiliated co. not in Country S, or a branch in Country S of X co.’s affiliated co. not in Country S, even if Country S cannot gather information in Country R. This idea, in which Country S’ taxing right on X co.’s income is not only based on direct nexus with X co.’s physical presence in Country S but also based on indirect nexus, can be considered to re- semble the idea in the 21

century which can be found in anti-fragmentation discussion con- cerning the scope of recognition of a PE. The most important choice in the era of the League of Nations was the adoption the idea of separate accounting, not adopting the {no taxation without a PE or an affiliated entity} rule and the force of attraction rule. Adoption of the idea of separate accounting now, in the existing international tax law system, provides scope in which USA’s multinational enterprises erode tax base through business restructuring while laughing at the efforts of countries who try to prohibit base erosion with strict applica- tion of the arm’s length principle.

Country R Country S

X co.---branch

X co.---X co.’s affiliated co.

X co.---X co.’s affiliated co.---The affiliated co.’s branch

Chapter III discusses what kind of derogation is needed concerning the arm’s length principle which is the backbone of the existing international tax law system.

Section III-1 explains the terms reliable arm’s length and unreliable arm’s length which are not common wordings.

The idea of separate accounting was an idea like labor value theory in which a person (individual or corporation) or a part (PE) that physically contributes to the business income is entitled to a share of the income according to its contribution (nowadays, expressed as value creation). However, in arm’s length transactions in the real world, a person who has contributed nothing concerning business income can earn income, as in a case in which the person earns compensation from a covenant not to compete. The legal validity of payments of compensation of covenants not to compete in arm’s length settings cannot be denied, but payments of compensation of covenants not to compete between affiliated entities could be denied. If we make a fiction of income earning according to the contribution, we must ac - cept the concepts of reliable arm’s length transactions and unreliable arm’s length transac- tions.

Section III-2 discusses mild derogations and severe derogations from the arm’s length

principle.

As mild derogations from the arm’s length principle, we will make a countermeasure in which income attribution recognized by unreliable arm’s length transactions even those ex- isting in the real world cannot be applied between affiliated entities. The idea of Amount B

which was recently discussed makes a fiction in which local activities must earn a certain minimum rate of return; this idea can be considered as ensuring income attribution accord- ing to the contribution by the local activities with the idea of labor value theory. And more, this idea can be considered as denying tax planning of base erosion with unreliable arm’s length transactions (income can be attributable to a person or a part according not to their physical contribution) in the country of the place of activities. If we understand the idea of Amount B as above, Amount B can be found to be a mild derogation from the arm’s length principle.

Section III-3 discusses severe derogations from the arm’s length principle.

The idea of Amount A

of taxing right allocation based on a place of demand cannot be justified by discussing relationship between reliable arm’s length transactions and unreliable arm’s length transactions. It is because Amount A discusses derogations from the traditional international tax law system in which demand alone does not give the base of geographical allocation of income. Therefore, discussion of Amount A can be found to be a severe dero- gation from the arm’s length principle. How can we justify such severe derogations? Sub- section III-3-2 tries to give a model, concerning the markets in which so-called network ef- fects have serious impacts, in which mild derogations from the arm’s length principle denying unreliable arm’s length transactions are not enough to convincingly accomplish taxing rights allocation among countries, and we will need change in thinking (severe dero- gations) that the geographic allocation of income can be based on demand in a market coun- try.

Chapter IV compares corporation tax and value added tax because taxing right allocation according to demand has traditionally been discussed in the context of value added tax.

II. Outline of existing international tax law system

II-1. Geographic allocation of income and personal attribution of income

Traditionally, when we make an image of a source of income

(geographical allocation

1

OECD/G20 Base Erosion and Profit Shifting Project: “Statement by the OECD/G20 Inclusive Framework on BEPS on the Two-Pillar Approach to Address the Tax Challenges Arising from the Digitalisation of the Economy”, As approved by the OECD/G20 Inclusive Framework on BEPS on 29-30 January 2020, p. 16 (2020 January 31 http://www.oecd.org/newsroom/in- ternational-community-renews-commitment-to-multilateral-efforts-toaddress-tax-challenges-from-digitalisation-of-the-econo- my.htm).

2

OECD, footnote 1, p. 9.

3

Author of this article usually criticizes the concept of global income, but {income} in this article is not concerned with the

difference between the concept of global income and consumption-type concept of income. And more, {income} in this article

does not necessarily differentiate between gross income (revenue) and net income (profit). {Profit} in this article clearly means

net income which is calculated as revenue minus cost.

of income), the standard of the source has long been considered as a place of assets or a place of business. However, as said below, this article states that we do not need the stan- dard of place of assets and we should look at only the standard of place of business. By the way, this article willfully uses an ambiguous expression of {image}, because geographical allocation of income cannot be deductively and logically determined

from the definition of income.

See six examples below.

(1) Country R’s X co. has an immovable property in Country S, which is rented to Y co., and gets rental income from Y co.

(2) Country R’s X co. contributes money to Country S’ Y co., and gets dividends income from Y co.

(3) Country R’s X co. lends money to Country S’ Y co., and gets interest income from Y co.

(4) Country R’s X co. has made invention in Country R, has got a patent right in Coun- try S, enters into license agreement of the patent right with Country S’ Y co., and gets royalty income from Y co.

(5) Country R’s X co. produces tangible properties, sells them to Country S’ Y co., and gets business income.

(6) Country R’s X co. supplies services to Country S’ Y co., and gets business income.

Country R Country S

(1) X co.<---rental income---Y co.

(2) X co.<---dividends income---Y co.

(3) X co.<---interest income---Y co.

(4) X co.<---royalty income---Y co.

(5) X co.<---sales income---Y co.

(6) X co.<---service income---Y co.

II-1-1. Rental income of immovable properties

In Example (1), most tax lawyers have an image that the geographical allocation of in-

4

Michael J. Graetz & Michael M. O’Hear, The “Original Intent” of U.S. International Taxation, 46 Duke Law Journal 1021, at 1034 (1997). Note that Graetz & O’Hear (1997) tried to support source tax jurisdiction.

5

Recently, the idea of LSR (location specific rent) attracts attention in order to justify special taxing rights such as DST (digi- tal service tax). This idea is attractive but this article does not discuss LSR nor DST. Fortunately, 渡辺徹也「デジタルサー

235頁 (2020) (Watanabe, Tetsuya, “Logical foundation and challenge of digital service tax: mainly concerning thoughts on Location-Specific Rent”, Financial Review, no. 143, pp. 219-235) considers LSR and DST. Cf. Wei Cui (

Services Tax on the Verge of Implementation”, 67:4 Canadian Tax Journal 1135-1152 (2019); Wei Cui, “The Superiority of the Digital Service Tax over Significant Digital Presence Proposals”, 72:4 National Tax Journal 839-856 (2019); Wei Cui, “The Digital Services Tax: A Conceptual Defense” (unpublished, 2019 April 30 https://ssrn.com/abstract=3273641); Wei Cui & Ni- gar Hashimzade, “The Digital Services Tax as a Tax on Location-Specific Rent” (unpublished, 2019 July 29 https://ssrn.com/

abstract=3321393); Daniel Shaviro, “Digital Service Taxes and the Broader Shift From Determining the Source of Income to

Taxing Location-Specific Rents” (unpublished, 2020 January 14 https://www.law.nyu.edu/sites/default/files/shaviro-singa-

pore-lecture-v3.pdf).

come from immovable properties are located in Country S

and an image that personal attri- bution of the income belongs to X co. (and in Country R). Most tax lawyers have long ac- cepted the fact that there can be naturally a discrepancy between geographical allocation of income and personal attribution of income.

I stress with the word {naturally} because, in the context of taxation on subsidiaries (in- cluding affiliated corporations) and PEs, there has been a tendency that the geographical al- location of income and personal attribution of income are merged (personal attribution of income is prior and geographical allocation of income is coordinated with personal attribu- tion of income); however, I would like to stress that the tendency is not natural. I will write about the tendency in section II-2, and this section continues to discuss geographical alloca- tion of income.

In Example (1), we can find several reasons of an image that the geographical allocation of income from immovable properties is located in Country S; first candidate of the standard concluding geographical allocation of income is the standard of place of assets in which source of income in the meaning of {from where} is considered to be the place of immov- able properties which is the source of income in the meaning of {from what}, which is lo- cated in Country S in Example (1); the second is the standard of place of business (differen- tiated from the third below, the standard of place of business of earners) in which source of income with the meaning of {from where} is considered to be the place of business (im- movable properties rent business in Example (1)) which is the source of income with the meaning of {from what}, which is located in Country S; third, is the standard of place of payers (differentiated from the second above, the standard of place of business of payers) in which source of income with the meaning of {from where} is the place of payers’ activities which is the source of income with the meaning of {from what}, which is located in Coun- try S. If the payer, Y co. is not a resident of Country S but of a third country (Country T), the third standard has a possibility to lead an image that the geographical allocation of income is located in Country T if the standard of place of payers looks at residency of the payer;

however we can ignore the possibility because no tax lawyers will adopt the standard of place of payers’ residency concerning immovable property rental income.

By the way, concerning rental income of immovable properties, there is an unignorable size of cases in which the standard of place of payers’ business is not suitable. When payers borrow immovable properties for their consumption (typically, in order to live in), the stan- dard is not suitable. Even though explaining geographical allocation of income from im- movable properties according to the standard of place of earners’ business or the standard of place of payers’ business would have not much difference, the standard of place of earners’

business would have less difficulties.

II-1-2. Dividends income

In Example (2) of dividends income, tax lawyers have little tendency to have an image

6

See OECD Model Tax Convention, Article 6.

that the money contribution itself is business of X co. (the contribution in Example (2) and the lending in Example (3) can be considered as a part of business of fund utilization, but traditionally the contribution itself has not been considered as business); it is common among tax lawyers that we have an image that the geographical allocation of dividends in- come is located in Country S because we look at the fact that the payer of the dividends is Y co. who is a resident of Country S (the standard of place of payers). However, pure form of the standard of place of payers (looking at only residence of the payers) has rarely, if any, been adopted; we have an image that the geographical allocation of dividends income is lo- cated in a third country (Country T) if the payer, Y co., does business activities and the busi- ness’s profit is the source ({from what}) of the dividends payment (the standard of place of payers’ business).

In Example (2) of dividends income, we can also explain the geographical allocation of the dividend income as existing in Country S because the source ({from what}) of the divi- dends is stocks and the stocks’ place is determined by looking at the residence of the issuer (Y co.) of the stocks, in short, Country S (the standard of place of asset). Derogating from the context of income taxation, stocks issued by a foreign corporation are not considered as a domestic asset in the context of inheritance tax and gift tax.

Returning to the context of income taxation, the standard of place of assets can be considered as the same as the stan- dard of place of payers’ residence concerning dividends income; however, the standard of place of payers’ business would be more persuasive than the standard of place of payers’

residence, so we do not have to adhere to the standard of place of asset.

II-1-3. Interest income

In Example (3) of interest income, as like in Example (2) of dividends income, it is pos- sible to explain that the geographical allocation of interest income is located in Country S because the loan claim is located in Country S (the standard of place of asset). However, as like in Example (2), we should not adhere to the standard of place of asset. And moreover, as like in Example (2), it is usual to explain that the geographical allocation of income is lo- cated in Country S because the payer of the interest, Y co., is located in Country S (the stan- dard of place of payers). And, as like in Example (2), we have an image of the geographical allocation of interest income, not relying on the pure form of the standard of place of payers’

residence, but relying on the standard of place of payers’ business because the source of in- come in the meaning of {from what} is the business activities of Y co.; if the source of {from what} is the business activities of Y co. located in third country (Country T), then the geographical allocation of the interest income is considered to be located in Country T.

However, Example (3) has points that differ from Example (2). Although it is unusual

7

In Takefuji case, Supreme Court, 2011 February 18

, reported in Hanrei Jihô, no. 2111, p. 3, the objects of the gift were

stocks (more precisely, 出資口数 (number of capital contributions)) of a Dutch corporation, but the corporation indirectly held

stocks of a Japanese corporation (Takefuji). An argument that the objects of the gift were domestic (Japanese) assets is not

strange when we look at economic situations in this case. However, the Japanese tax authority did not argue that the objects of

the gift were domestic asset because such argument would have little chance in law.

that the money contribution in Example (2) is considered as business activities of X co., the lending of money in Example (3) can be considered as business activities of X co. Let’s sup- pose that X co. has borrowed money (¥8000) from a third party (Z co.) with interest rate of 10% and lends the money (¥8000) to Y co. with interest rate of 12.5%; X co. gets interest income of ¥1000 (gross income) from Y co., pays interest income of ¥800 to Z co., and gets net income (or profit) of ¥200 (ignoring other cost for simplicity). If X co. has business fa- cilities in only Country R, it is possible to have an image that the geographical allocation of the business profit of ¥200 is located in Country R (the standard of place of earners’ busi- ness). However, on the one hand, when X co. does not have a PE in Country S, it has been rarely seen to have an image that the geographical allocation of interest income (regardless whether net income or gross income) is determined by the standard of place of earners’ busi- ness.

On the other hand, when X co. has a PE in Country S, it is more widely accepted to have an image that the geographical allocation of the business profit is located in Country S by the standard of place of earners’ business. Here, business profit means net income. And what is more, when X co. has a PE in Country S, also dividends income is included into the con- text of taxation on profits attributable to the PE.

II-1-4. Royalty income

In Example (4) of royalty income, the typical image is that the geographical allocation of the royalty income concerning license of Country S’ patent right is located in Country S.

We have to note that many existing tax treaties have the {no taxation without a PE} rule concerning royalty income

; there is a complication that the geographical allocation of the royalty income is located in Country S but Country S’ taxing right is prohibited.

Even with this complication, there seems to be little space to argue that the geographical allocation of the royalty income is not located in Country S.

Why do we have such an image of the geographical allocation of royalty income? It is because Country S’ patent right is definitely located in Country S, in the light of the idea of the principle of territoriality in intellectual property law textbooks.

In this regard, tax professionals sometimes say that intangible properties can be easily transferred across borders internationally; however, not concerning all intangible properties but, at least, concerning intellectual properties like patent rights, we should think that the rights cannot be transferred across borders as well as immovable properties.

8

See OECD Model Tax Convention, Article 10(4) (PE taxation on dividends income) and Article 11(4) (PE taxation on inter- est income).

9

Compare OECD Model Tax Convention, Article 12(1) (allocating taxing rights on royalty income only to Country R in ex- ample (4)) and Articles 10(2) and 11(2) (allocating limited taxing rights on dividends income and interest income to Country S). Also see UN Model Tax Convention, Article 12(2) (allocating limited taxing rights on royalty income to Country S).

10

In many countries, domestic tax law allocates taxing rights on royalty income to Country S in Example (4) if the country’s taxing rights are not restricted by a tax treaty.

11

Paris Convention for the Protection of Industrial Property, Article 4 bis (Independence of Patents Obtained for the Same In-

vention in Different Countries) ( https://wipolex.wipo.int/en/text/287556 ); See, 田村善之『知的財産法　第2版』444頁 （有

(Tamura, Yoshiyuki, “Intellectual Property Law second edition”, p. 444, Yûhikaku, 2000).

If it is right statement that intangible properties can be easily transferred across borders, there can be two possibilities.

The first possibility is that although patent rights are governed by the Paris Convention’s principle of territoriality, there can be a space in which not all intangible properties are gov- erned by the principle of territoriality. However, I have not yet found such statement that in- tellectual properties are governed by the principle of territoriality but other types of intangi- ble properties can be easily transferred across borders. In international tax law discussions, the difference of transferability across borders between intellectual properties and other types of intangible properties has not gathered attraction. Therefore, this article ignores this first possibility.

The second possibility is that although patent rights or other intellectual properties can- not be transferred across borders like immovable properties, it is possible for buying agree- ments of certain immovable properties to cross borders; therefore, it is possible to see that there can be a transfer across borders of an owner of a right, regardless whether a right of an intellectual property or of an immovable property, and the residence country of the owner of the right can be changed. The second possibility is not wrong. But, have we explained that immovable properties can be easily transferred across borders? If not, we should refrain from explaining that intangible properties can be easily transferred across borders.

As said above in two paragraphs, this article does not say that intangible properties can be easily transferred across borders.

The image becomes less robust as we can see that the geographical allocation of the roy- alty income is located in Country S according to the standard of place of assets in Example (4), which differs when compared with Example (1) of the rental income of the immovable property. On the one hand in Example (1), it can be easily thought that the business of rental of the immovable property is done in Country S; on the other hand in Example (4), the in- vention which is base of the royalty income has been done in Country R. If we apply the standard of place of business in Example (4) as well as in Example (1), the geographical al- location of the royalty income should be considered as be located in Country R rather than Country S.

In this regard, readers might have an image that the geographical allocation of the royal-

ty income is located in Country S not relying on the standard of place of earners’ business as

in Example (1) but relying on the standard of place of payers’ business as in Examples (2)

and (3), because the royalty income is paid out of the fruits of the business activities of Y

co. in Country S. Also in application of Japanese domestic law, we can find examples that,

in deciding the scope of royalty income which is subject to withholding tax, people might

have an assumption that income is paid out of the fruits of business activities of a domestic

payer not relying on the standard of place of earners’ business but relying on the standard of

payers’ business.

However, if we can ignore the historical background, and if we can make

an assumption that the royalty income is compensation of X co.’s invention in Example (4)

(subsection III-2-3 will discuss an example in which this assumption is relaxed), it is strange

that the geographical allocation of the royalty income is governed by the place of payers’

business as in Examples (2) and (3). OECD Model Tax Convention, Article 12 treats royalty income as located between Articles 10 (dividends income) and 11 (interest income) and Ar- ticle 13 (capital gain), and this location might give us an impression that royalty income is a kind of capital income; however, if we care about the distinction between financial transac- tions and real economic transactions,

and if it is assumed that dividends income and inter- est income are earned in financial transactions and royalty income is compensation of inven- tions, then we should look the invention which is real economic activity (operation of real production factors such as humans, machines, or factories) and the royalty income should be considered as earnings in the context of real economic transactions.

Between the standard of place of assets or the standard of place of payers’ business and the standard of place of earners’ business, we have different theories of the geographical al- location of the royalty income. This difference gives us an intuition that our image of the geographical allocation of the royalty income is likely to be a watershed among some stan- dards of source of income of {from where}.

This article argues that the standard of place of assets can be absorbed into the standard of place of payers’ business. As discussed in Example (2) of dividends income, readers might see that the place of the stocks is considered to be the residence of the issuer, Y co.;

however, when the business activities of Y co. which is the base of the dividend income are engaged in mainly third country (Country T), the standard of place of assets would give a counter-intuitive result concerning the image of the geographical allocation of income. In Example (1) of the immovable property, the standard of place of assets seems to fit with an image of the geographical allocation of income in a most persuasive way, but we do not need to adhere to the standard of place of assets because we can reach the same result if we apply the standard of place of earners’ business and the standard of place of payers’ busi- ness. Therefore, abolition of the standard of place of assets would lead to few mistakes.

In Example (4) in contrast to Example (1), the standard of place of payers’ business and the standard of place of earners’ business leads different results of images of the geographi- cal allocation of income. We need to think carefully. I said {if we can ignore the historical background, and if we can have an assumption that the royalty income is compensation of X co.’s invention in Example (4)}, and this article argues that the standard of place of earners’

business is suitable for the royalty income and that the geographical allocation of the royalty income should be considered to be located in Country R, not in Country S. This argument is different from a traditional image that the geographical allocation of royalty income has

12

See,

弘先生還暦記念論文集　知的財産法の理論と現代的課題』580頁 （弘文堂，2005） (Asatsuma, Akiyuki, “Scope and al-

location concerning royalty income of intellectual properties”, Aizawa, Hidetaka = Ôbuchi, Tetsuya = Koizumi, Naoki = Ta- mura, Yoshiyuki, ed., In Honor of 60th Birthday of Professor Nakayama, Nobuhiro: Theory and current challenge in Intellectual Property Law, p. 580 (Kôbundô, 2005)).

13

Distinction between financial transactions and real economic transactions can be difficult in, for example, rental transactions

of machines. See footnote 39. Concerning cash flow tax with R base or R+F base, see, 神山弘行「法人課税とリスク」金子

(Kôyama, Hiroyuki, “Corporation

taxation and risk”, Kaneko, Hiroshi = Nakazato, Minoru = J. Mark Ramseyer, ed., Tax Law and Market, p. 321, Yûhikaku,

2014).

been located in Country S (regardless whether relying on the standard of place of assets or the standard of place of payers’ business).

This difference gives us serious challenge concerning how we have an image of the geo- graphical allocation of income. This article repeatedly and intentionally uses the word of {image} which is ambiguous. It is because the geographical allocation of income cannot be decided by logic. {Image} and {should} are incompatible. The argument that the geographi- cal allocation of income is only an {image}, not decided by logic, requires us to take anoth- er step concerning the geographical allocation of income: even if the source rule of royalty income and the source rules of immovable properties income, dividends income, or interest income lead to different images of the geographical allocation of income, we should try to sublate (aufheben) the different images. However, I have not been able to sublate them and no one seems to have.

This difficulty of sublating the different images makes me to argue that the traditional image that the geographical allocation of the royalty income in Example (4) is located in Country S, is incompatible with images of the geographical allocation of other income in Examples (1) (2) (3) (5) and (6). If we can have the assumption that the royalty income is compensation of X co.’s invention in Example (4) (but see, subsection III-2-3), the geo- graphical allocation of the royalty income should be considered to be located in Country R with the standard of place of earners’ business.

II-1-5. Sales income

In Example (5) of sales income, if X co. has a PE in Country S, the geographical alloca - tion of profit in Country S is limited to profit attributable to the PE.

If X co. does not have a PE in Country S, a leading case

of sales income is concerned with the title passage rule.

However, the title passage rule is a proxy for finding the place of business and the proxy is not well-tuned. Roughly speaking, if X co. does not have a PE in Country S, we have an im- age that the geographical allocation and the personal attribution of the sales income is locat- ed in Country R.

II-1-6. Service income

Example (6) of service income is easy to think of like Example (5). If X co. has a PE in Country S, OECD Model Tax Convention, Article 7(1)’s second sentence is applied. X co.

does not have a PE in Country S in a leading case,

a Mexican radio broadcasting enter- prise’s case. In this case, the Mexican enterprise got advertising income from American en- terprises but it was ruled that the Mexican enterprise was not engaged in trade or business in the USA. (By the way, some tax lawyers might discuss that whether the Mexican enterprise had a PE in the USA or not. I think that this discussion has little meaning because even if the PE was found, the profit attributable to the PE would be small.

) Roughly speaking, if X

14

OECD Model Tax Convention, Article 7(1) second sentence and also, paragraph (2).

15

U.S. v. Balanovski, 236 F.2d 298 (2

Cir. 1956); reversing 131 F.Supp. 898 (1955).

16

Commissioner v. Piedras Negras Broadcasting Co.,127 F.2d 260 (5

Cir. 1942), affirming 43 BTA 297 (1941).

co. does not have a PE in Country S, we have an image that the geographical allocation and the personal attribution of the service income is located in Country R.

II-1-7. The standard of place of payer ≠ the standard of place of demand

In Examples (1) (2) (3) and (4), it is not impossible to explain that the geographical allo- cation of rental income, dividends income, interest income and royalty income is decided by the standard of place of demand.

However, in Examples (5) and (6) of business income, the standard of place of demand is hard to apply. This article tries to find a more compatible explanation about images of the geographical allocation of several types of income (although impossible concerning Exam- ple (4)), and therefore, the standard of place of demand is denied.

If, in Example (1), the geographical allocation of rental income can be explained by the standard of place of earners’ business, and if, in Examples (2) (3), the geographical alloca- tion of dividends income and interest income can be explained by the standard of place of payers’ business, we can rely on the standard of place of business (although incompatibility exists between earners’ business and payers’ business) concerning Examples (1) (2) (3) (5) and (6).

In Examples (2) and (3), if we do not look at X co.’s business profit (net income) but at dividends or interest payment by Y co. to X co. (gross income) and if we think that X co.

only does financial transactions and payments of dividends or interest income are parts of the value added produced by Y co.’s real economic activities (that is, operations of real pro- duction factors such as humans, machines, or factories), Examples (1) (2) (3) (5) and (6) can be compatibly explained that the geographical allocation of income has been imaged look- ing at the place of real economic activities producing value added. This article tries to sub- late (aufheben) the standard of place of earners’ business in Examples (1) (5) and (6) and the standard of place of payers’ business in Examples (2) and (3) as the standard of place of production.

II-1-8. Rethinking about royalty income: protection by governments is not easy to explain the geographical allocation

In Examples (1) (2) (3) (5) and (6), we can explain the geographical allocation of in- come according to the standard of place of production; on the other hand, is an image of the geographical allocation of the royalty income in Example (4) incompatible?

In Example (4), some tax lawyers might think that because Country S’ government gives protection in intellectual property law (patent law, copyright law, or etc.), the geographical allocation of the royalty income should be considered to be located in Country S, and that it is not persuasive that the standard of place of production is compatible in Examples (1) (2) (3) (5) and (6) and is also suitable in Example (4). However, this article argues that protec-

17

See,

(Oka,

Naoki, “Implications of ‘new taxing rights’ and ‘minimum tax’ propositions and prospects for international tax law”, The Japa-

nese Association of International Business Law Academic Journal, no. 5, pp. 61-86 (2020)).

tion by governments is not a reason, because, in Examples (5) and (6), the geographical al- location of business income is considered to be located in Country R even though X co. gets business income with help of Country S’ legal system.

In order to explain that the geographical allocation of the royalty income in Example (4) is located in Country S in compatible way with images that the geographical allocation of the business income in Examples (5) and (6), it can be possible to explain that X co.’s bene- fit derived from the legal system in Country S is different between Example (4) and Exam- ples (5) and (6). However, this article argues that this explanation is not persuasive. Even if Country S does not have intellectual property law such as patent law or copyright law, it is not certain that X co.’s invention or creation is free to utilize in Country S. For example, Ja- pan has a precedent case in which B co.’s imitation of wood grain papers constitutes illegal act against A co. and B co. must pay damages to A co. (Civil Code, Article 709),

even though A co.’s wood grain papers are not protected by copyright law.

Even though the wood grain papers are not protected by copyright law, we think that the wood grain papers are intellectual properties of A co. It is not impossible to explain that the intellectual proper- ty law’s protection and its extent in Country S give a line-drawing concerning the geograph- ical allocation of income in Example (4) and in Examples (5) and (6) because legal protec- tion in Example (4) is stronger than in Examples (5) and (6), but I argue that this explanation is not compatible with Examples (2) and (3) because we find little difference of strongness of legal protection between Examples (5) and (6) and Examples (2) and (3).

If the intellectual property law’s protection and its extent in Country S give a line-draw- ing concerning the geographical allocation of income in Example (4) and in Examples (5) and (6), why does the legal system protecting the dividend income and the interest income in Examples (2) and (3) justify images that the geographical allocation of income of the div- idend income and the interest income is located in Country S? If intellectual property law’s protection is special, the geographical allocation in Country S in Examples (2) and (3) would be denied. It is not impossible to explain that protection by Country S’ government justifies images that the geographical allocation of income is located in Country S in Exam- ples (2) (3) and (4) and does not justify in Examples (5) and (6); however, this explanation is not an explanation but a conclusion.

When trying to make compatible explanations about images of the geographical alloca- tion of income, I argue that it is persuasive to explain that the standard of place of produc- tion has been adopted even though this explanation is not compatible in Example (4), be- cause the incompatibility is small.

In my view, the standard of place of protection by governments is not good explanation.

If the protection by government justify taxing rights, such tax would be based on compensa- tion theory. Compensation is discussed in the context of efficiency. In the light of efficiency, Country S can impose tax not exceeding the extent of benefit of protection by Country S’

18

Today, the damages have legal base in Unfair Competition Prevention Act.

19

Wood grain paper case, Tokyo High Court, 1991 December 17

, reported in Hanrei Jihô, no. 1418, p. 120.

government. However, income tax is not only discussed in the context of efficiency but also in the context of equity or distributive justice, because income tax has the intent of redistrib- uting income.

In order to reduce the income gap, the government must impose tax exceed- ing the extent of benefit of protection by the government.

II-1-9. Choice between the standard of place of demand or the standard of place of production is not decided with logic

As I said repeatedly, the geographical allocation of income is not decided by logic, and is only an image which tax lawyers have had historically. Therefore, the explanation that the standard of place of production can be compatible with more types of income than other standards only belongs to a sein discussion (a discussion about how it is), not a sollen dis- cussion (a discussion about how it should be).

Source of income of {from what} is value added.

Value added is only a calculation of revenue minus purchase. The geographical allocation of value added also should be dis- cussed as well as the geographical allocation of income. If we go to the next step, that is, a sollen discussion, possible alternatives are the place of demand or the place of production.

In a sein discussion, traditionally the geographical allocation of income has been accord- ing to the standard of place of production; this sein discussion does not necessarily give a base of disadvantage of the standard of place of demand in a sollen discussion. Since the end of the 20

century,

academic tax scholars started to discuss about the standard of place of demand in the light of electronic commerce

(although incompatible with the traditional geographical allocation of income in the context of a sein discussion), and such discussion is not strange in a sollen discussion. It is because there is no logical answer whether source of income of {from what} is production or demand.

In the 21

century, the wording of {value creation} is replaced with the traditional word- ing of source of income. Source of income (geographical allocation of income) is tradition- ally imaged by the standard of place of production; it might be guessed that the wording of value creation produces (is chosen in order to produce) a space in which we can play of tug- of-war about taxing rights allocation in the 21

century between the standard of place of production and the standard of place of demand.

II-2. Choice since the era of the League of Nations: (1) no taxation without a PE (2) separate accounting (3) attributed income principle

When Country S tries to impose tax on business income of a foreign enterprise (X co.)

20

Some lawyers hate to use the word {redistribution} but this article does not address distribution and redistribution. See, Liam Murphy & Thomas Nagel, “The Myth Of Ownership: Taxes and Justice” (Oxford University Press, 2002).

21

In this article, I do not address the global income concept and consumption-type concept of income (footnote 3), and do not address income-type of value added and consumption-type of value added.

22

See, Unknown, “Disappearing Taxes”, The Economist, 1997 May 31

, p. 19.

23

See, Richard L. Doernberg, “Electronic Commerce and International Tax Sharing”, 16 Tax Notes International 1013-1022

(1998 March 30); Richard L. Doernberg & Luc Hinnekens, Electronic Commerce and International Taxation (Kluwer, 1999).

(ignoring withholding tax on capital income), there are three keys. The first is threshold. The second is income allocation between a domestic entity and foreign entities. The third is the scope of income which is subject to tax filing system. About these three keys, the left sides are the traditional choice since the era of the League of Nations to the OECD Model Tax Convention and right sides are alternatives.

(1) <Threshold>

No taxation without a PE vs. No taxation without a PE or an affiliated entity (2) <Income allocation>

Separate accounting + arm’s length principle vs. consolidation + formulary apportion- ment

(3) <Scope of income filed>

Attributed income principle vs. entire income principle (force of attraction)

Traditionally, limitation on tax on a foreign enterprise looks at (1) threshold at first, that is {no taxation without a PE} rule (subsection II-2-2 will explain {no taxation without a PE or an affiliated entity} rule). In negotiation between Country R and Country S when entering into a tax treaty, the width of the definition of a PE is easy to negotiate (OECD and UN Model Tax Convention, Article 5). On the other hand, concerning (2) income allocation, OECD member countries have had a consensus that the arm’s length principle is used gen- erally and formulary apportionment can be used only exceptionally (nonmember countries might be reluctant to agree with this consensus but they also know the consensus).

However, in this article, I argue that the most epoch-making choice in the era of the League of Nations when the backbone of the existing international tax law system was shaped is (2) income allocation; at that time, the idea of separate accounting was adopted.

The wording of {epoch-making choice} has two meanings; the first is that the choice has af- fected international tax law system broadly; the second is that the choice was not an {un- avoidable choice} at that time in the light of limitation of enforcement capacity in Country S.

But, first, (1) threshold is discussed.

II-2-1. (1) Threshold (no taxation without a PE)

In the light of limitation of enforcement capacity in the first half of the 20

century (without international exchange of information as in the 21

century), when Country S tries to impose tax on business income of a foreign enterprise (X co.), Country S cannot easily reach its taxing arms to X co. without X co.’s physical presence in Country S; therefore, choice of {no taxation without a PE} rule was, if not unavoidable (because withholding tax not only on capital income but also on business income is not impossible), a nearly unavoid- able choice. The OECD Model Tax Convention adopted a threshold of PE concept and American domestic tax law historically adopted a threshold of {trade or business in the USA} concept; some tax lawyers might think that PE concept was not unavoidable choice.

However, there is little difference between the PE concept and the {trade or business in the

USA} concept. When a foreign enterprise has no physical presence in the USA, {trade or business in the USA} threshold is hardly accomplished. Regardless whether OECD Model or American domestic law (or other countries’ domestic law), some physical presence of a foreign enterprise is required. Japanese domestic tax law adopted an exception of {no taxa- tion without a PE} rule concerning personal service income, in which service income of a foreign resident can be taxed in Japan even if he/she has no PE in Japan; however, this rule requires real performance of services in Japan and it also can be explained that physical presence of his/her body is required to justify Japanese taxing right.

In the context of VAT (value added tax), discussion about taxation on a foreign enterprise without physical pres- ence started at the end of the 20

century, but, regardless whether in the context of VAT or income taxation, physical presence was required in order to enforce taxing power against a foreign enterprise until the end of the 20

century.

Nowadays, a service PE concept

is discussed, but Japanese domestic tax law’s imposi- tion of tax on personal service income without an establishment functionally resembles a service PE taxation. Taxation on entertainers and sportspersons by OECD Model Tax Con- vention, Article 17 was categorized as an exception of the PE concept in the 20

century, but in the 21

century, Article 17 can be categorized as one example of the service PE concept.

If we do not care about what we call {service PE}, {personal service income taxation with- out a PE in Japan}, or {OECD Model, Article 17}, it can be said that the key of {no taxation without a PE} rule requires a domestic physical presence of a foreign enterprise, not the definition of the PE concept.

And moreover, I said {withholding tax not only on capital income but also on business income is not impossible}, and in reality, UN Model Tax Convention, Article 12(2) or other domestic tax laws impose withholding tax on royalty income. It can be said that the key is what types of income can be categorized as UN Model, Article 12(2). India is famous (infa- mous?) about this kind of withholding tax on fees for technical services, but India is not an exception. Japan also tried to widen objects of withholding tax in the 20

century (see, foot- note 12).

In summary, concerning (1) threshold in the light of limitation of enforcement capacity in first half of the 20

century, Country S must have relied on withholding tax or domestic physical presence when trying to impose tax on a foreign enterprise; that is, nearly unavoid- able choice. However, as discussed later in subsection II-2-2, {relying on domestic physical presence} does not necessarily mean the {no taxation without a PE} rule. The {no taxation without a PE} rule is a subcategory of {relying on domestic physical presence} and there can be another subcategory, that is {no taxation without a PE or an affiliated entity} rule.

24

About Income Tax Act, Article 164(1)(iv), see ASATSUMA, Akiyuki, “Japan”, in IFA 2012 Boston Congress: cahiers de droit fiscal international, volume 97a, Enterprise services, pp. 413-435 (2012).

25

This article does not address (old) OECD Model Tax Convention, Article 14’s “fixed base” concept because the fixed base

concept resembles the PE concept.

II-2-2. If, concerning (2) income allocation, we do not adopt separate accounting, is there an alternative concerning (1) threshold?

Concerning (2) income allocation, the League of Nations (especially, Carrol) adopted the idea of separate accounting, as explained by Fuchi, Keigo or other tax law scholars.

This article does not assess history. In order to look at functions of international tax law system, this article thinks a space in which the international tax law system could have ad- opted an idea other than separate accounting concerning (2) income allocation.

A counter part of separate accounting is consolidation. In the international tax law sys- tem, separate accounting fits well with the arm’s length principle and consolidation fits well with formulary apportionment. Therefore, traditionally international tax law discussions mainly look at the choice between the arm’s length principle vs. formulary apportionment rather than separate accounting vs. consolidation. {Vs.} sounds like equal chance of the arm’s length principle and formulary apportionment, but, actually, the arm’s length principle is considered general and formulary apportionment is considered an exception.

If we do not adhere to separate accounting, what kind of international tax law system could have been evolved?

Regardless whether we adopted separate accounting or not, we have little room concern- ing (1) threshold issue; we must rely on withholding tax or rely on domestic physical pres- ence in first half of the 20

century. Concerning {relying on domestic physical presence}, tax lawyers have long discussed the width of the PE concept. However, if we do not adhere to separate accounting, the widening of the PE concept is not only one way that Country S can lengthen its taxing arms to a foreign enterprise. Another way is {no taxation without a PE or an affiliated entity} rule.

Let’s see Examples (1) – (6) in section II-1. Country R’s X co. has entered into Country S in some economic sense and we discuss whether Country S can impose tax on Country R’s X co. or not. I confirmed that demand itself in Country S is hardly a base of the geographi - cal allocation of income in Country S, although an exception is Example (4) with royalty in- come. The width of the PE concept looks at the extent of presence of X co.’s direct alter ego in Country S. However, in real international transactions, physical presence in Country S is not only X co.’s direct alter ego (typically, a branch), but also X co.’s indirect affiliated enti - ty. If we adhere to separate accounting, X co.’s affiliated co. established in Country S is not a PE of X co. in Example (b) (see below), and a branch in Country S of X co.’s affiliated co.

not established in Country S is not a PE of X co. in Example (c). In Examples (b) and (c), the physical presence in Country S has a possibility to become an agent PE of X co., but even if an agent PE is recognized, profits attributable to the PE would be very small (profits attributable to an agent PE is discussed also in subsection III-2-2).

If we do not adhere to the idea of separate accounting, Country S’ domestic physical presence not only in Example (a) but also in Examples (b) and (c) can be help for enforcement of Country S’ tax on X co.

26

See,

(Fuchi, Keigo, “International Aspects of Income Taxation”

Yûhikaku, 2016).

It can be said that not only a direct PE of X co. in Example (a) but also an indirect affiliated entity of X co. in Examples (b) and (c) are the threshold; we can call the rule the {no taxa- tion without a PE or an affiliated entity} rule rather than the {no taxation without a PE} rule.

Country R Country S

(a) X co.---branch

(b) X co.---X co.’s affiliated co.

(c) X co.---X co.’s affiliated co.---The affiliated co.’s branch

Let’s think reversely. Concerning (1) threshold in the light of limitation of enforcement capacity in first half of the 20

century, {relying on domestic physical presence} does not exclude the possibility to adopt {no taxation without a PE or an affiliated entity} rule; there- fore, the reason that we did not adopt the {no without taxation without a PE or an affiliated entity} rule does not exist in the context of (1) threshold but in the context of (2) income al- location, in which, in real history, we adopted the idea of separate accounting. Logically, the {no taxation without a PE} rule in (1) threshold discussion is not the root of adopting sepa- rate accounting in (2) the income allocation rule, because the {no taxation without a PE}

rule can be enforced with other income allocation rules than separate accounting (that is, formulary apportionment). Therefore, I argue that the choice in (2) income allocation was more epoch-making than the choice in (1) threshold.

The {no taxation without a PE or an affiliated entity} rule was not adopted in real histo- ry. However, we can find a similar attitude in an Italian case,

in which a German corpora- tion was considered to have an agent PE with the existence of the affiliated corporation of the German corporation in Italy. If we care about the real history, a PE in Italy can be found by direct link between Italian physical presence and the German corporation; indirect link between Italian affiliated corporation and the German corporation is not a base to find a PE.

At that time, Italian Corte Suprema Di Cassazione was blamed by international tax lawyers (also Italian lawyers). This case shows that if the idea of separate accounting was applied rightly, Italian taxing rights would be seriously restricted. The Italian court gave taxing pow- er to Italy. And roughly ten years later, OECD discussed anti-BEPS measures and, in discus- sion of widening of PE concept, OECD adopted anti-fragmentation measures

which try to

27

There has been discussion about calculation of profits attributable to an agent PE; single taxpayer approach and double tax- payer approach (or dual taxpayer approach). OECD has not officially adopted single taxpayer approach with fear for losing face. However, when we adhere to the arm’s length principle, even if we assume double taxpayer approach, profits attributable to an agent PE is not much different than if we assume a single taxpayer approach. Many international tax lawyers who can speak without being constrained by their position think that the double taxpayer approach is near death. Even if the double tax- payer approach has meanings, Country S’ tax base would barely be substantially increased.

28

Ministry of Finance (Tax Office) v. Philip Morris GmbH, Corte Suprema Di Cassazione (Sezione Tributaria), Rome, Judg- ment date: 2002 March 7

, reported in 4 International Tax Law Reports 903-946 (2002) (concerning VAT, 947-1008).

29

See, OECD/G20, “Preventing the Artificial Avoidance of Permanent Establishment Status, Action 7 - 2015 Final Report”

(2015 October 5 https://doi.org/10.1787/9789264241220-en) p. 39, “Fragmentation of activities between closely related par-

ties”. Also see, p. 42, “Splitting-up contracts”.

find a PE with a lack of respect with the historical adherence to a direct link between domes- tic physical presence and a foreign enterprise even when a foreign enterprise tries to seg- mentalize activities in order to avoid a PE recognition. I understand that international tax lawyers now withdraw the blame against the Italian court.

II-2-3. Choice in the era of the League of Nations: the idea of labor value theory?

Subsection II-2-1 discussed that concerning (1) threshold, withholding tax or relying do- mestic physical presence are affordable way in the era of the first half of the 20

century.

Choice in the era of the League of Nations was that concerning income (dividends, in- terest, royalty, or etc.) subject to withholding tax, geographical allocation was given more respect than personal attribution (although I argue that the geographical allocation of royalty income is not compatible with other types of income in subsection II-1-4).

On the other hand, the {no taxation without a PE} rule and the attributed income princi- ple (OECD Model, Article 7(1) first and second sentences) put more respect on personal at- tribution than geographical allocation. PE taxation is a matter of Country S’ source tax juris- diction, but weakens the idea of geographical allocation.

If we do not adhere to separate accounting in the context of (2) income allocation, the {no taxation without a PE or an affiliated entity} rule concerning (1) threshold can be adopt- ed even in the light of limitation of enforcement capacity in the first half of the 20

century, and in the context of (3) scope of income subject to filing not subject to withholding taxa- tion, the attributed income principle is not necessary and the entire income principle (adopt- ed before the 1966 amendment in the USA and before the 2014 amendment in Japan) with the force of attraction rule would be subject to less objection than in real history.

Even though it is possible to adopt the entire income principle if we do not adhere to separate accounting, the geographical allocation of business income might be adhered to the standard of place of production (see subsection II-1-9) and even the entire income principle cannot substantially increase the tax base in Country S (ignoring dividends income, interest income, royalty income, or etc. which are subject to withholding tax). On the other hand, if the standard of place of demand is applied concerning the geographical allocation of busi- ness income, the entire income principle can substantially increase the tax base in Country S.

I argue that adopting separate accounting was an epoch-making choice (see subsection II-2-2), but a choice between the standard of place of production and the standard of place of demand concerning geographical allocation of income also has affected the international tax law system broadly.

The standard of place of production might have seemed to be more natural than the stan- dard of place of demand because income taxation is taxation on income earners in the first half of the 20

century. Therefore, I do not call the choice of the standard of place of pro- duction as an epoch-making choice. However, the broadness of impacts from the choice be- tween the standard of place of production and the standard of place of demand might be as large as the choice of adopting separate accounting.

Why did we adopt the standard of place of production rather than the standard of place

of demand and why did we place more respect on personal attribution of income rather than geographical allocation of income? I do not have enough evidence but I guess that the inter- national tax law system since the era of the League of Nations has been based on labor theo- ry of value (note that this sentence belongs to a sein discussion, not a sollen discussion).

Original labor theory of value looks at conflicts between laborers and capitalists who have machines or other real production factors other than human capital; but this article’s atten- tion to labor theory of value explains that source of income in the meaning of {from what}

is an operation of real production factors (humans, machines, factories, or etc.) and draws no line between humans and machines, factories, or etc. Labor theory of value will explain that source of income of {from where} is determined the location of source of income of {from what}, which means real production factors.

If labor theory of value has been the thoroughbass (basso continuo) of international tax law system since the era of the League of Nations, it is natural that discussions about the arm’s length principle in the 21

century have tendency to look at contribution of human capital and tendency to disrespect unreal contribution from a so-called cash box corporation.

And more, GILTI in American domestic law also respects operations of real production fac- tors and is compatible with labor theory of value.

III. Mild and severe derogation from the arm’s length principle III-1. Reliable and unreliable arm’s length transactions

The choices since the era of the League of Nations are, concerning (1) threshold, {no taxation without a PE} rule, concerning (2) income allocation, separate accounting, and con- cerning (3) scope of income subject to filing, and attributed income principle; however, in some cases, a real arm’s length transaction which is treated as a comparable transaction in a transfer pricing case has a possibility to lead to absurd results even if we believe in the arm’s length principle.

III-1-1. Compensation for non-action: Korfund case

A leading case

about compensation for non-action treats source of income derived from a covenant not to compete. In this case, a German corporation, Zorn co. had been obli- gated not to compete in North America and received payments from American corporation, Korfund co.

Zorn co. and Korfund co. were not unaffiliated enterprises, but they fought the amount of the compensation not to compete seriously. Therefore, we can say that the amount was

30

In my narrow search, Itai Grinberg, “International Taxation in an Era of Digital Disruption: Analyzing the Current Debate”, Taxes (2019 March) pp. 85-118, at 89 is the first article which refers to Karl Marx. Also see, Itai Grinberg, Stabilizing ‘Pillar One’: Corporate Profit Reallocation in an Uncertain Environment (unpublished, 2019 July 26 https://ssrn.com/abstract=3429863).

I have said Karl Marx orally, but I did not have the bravery to write Karl Marx in my old articles.

31

Korfund v. Commissioner, 1 T.C. 1180 (1943). Although the issue was the source of income and the result was that the

source of income existed in the USA, this article looks at deductibility of the payment, which was not the issue in the real case.